A Comparative Study Of Texture Features For Image Segmentation

Image segmentation is one of the most significant tasks in image processing. The outcome of image segmentation is a group of regions that collectively cover the entire image, each of the pixels in a region are homogeneous with respect to some characteristic or computed property, such as color, intensity, or texture. Already there are many approaches proposed for texture feature extraction which can be useful for image segmentation. One of the important issues here is how well these methods work on differentiating various textures that are available in a single image. This paper considers two texture measures namely Texture Spectrum and Uniform Local Binary Pattern for texture segmentation and evaluates their performance based on the segmentation accuracy. Two different synthetic images are used in experiments. One image contains four different textures and another one contains two different textures. MATLAB has been used for the implementation purpose.